I will use this thread to help brainstorm for a RSI based strategy that I am building to enter in the "battle of the bots" on this site. Over the past few weeks I have started exploring and documenting which strategies work and don't work with RSI. I plan to eventually document all of the indicators, but for now I will start with RSI to get a strategy up and running. Through that research I have found a few uses for RSI that seem to provide an edge.

So far the two most promising things I have found is:
#1 Buy when RSI is below a certain threshold. I have optimized on this strategy to find the best values. RSI(3) < 30 does well.
#2 Using Bullish Hidden Divergence.

Both of these methods seem to provide an edge that exceeds what you would expect due to random chance. If you want to see the full details of my RSI research, you can find it on my blog https://thestrategictrader.com

My blog is free and does not contain any advertisements. Its meant as an easy way to share multiple images and files as well as document my research. All of the source code used for my research is posted there as well and I will be referencing posts there throughout this process to keep everyone on the same page. If its inappropriate to post a link to my blog, please remove it.

I will continue to document and post my research on my blog throughout this process, but I think this forum may be well suited for exploring and kicking around strategy ideas on how to use what I have already found to produce a profitable strategy. There are only a few days left to get a submittal in for March so I need to work overtime.

Right now most of my research uses a 5 bar pre-defined exit. This is so that when comparing strategies exits do not affect the outcome. It isolates the entries so that I can find meaningful ones.

Any thoughts?

Thanks for sharing the code... I've never seen divergence work with any oscillator at all.